Malignancy, defined as the ability of a growing tumor to kill its host, lies at the root of the cancer problem. The mechanisms underlying malignant lethality remain mysterious and can benefit from analysis in simple, reductionist model systems. Many features of malignancy are seen in the `neoplastic' tumors of Drosophila, which are overproliferating, disorganized and dedifferentiated cell masses that kill hosts and also show other tumor-host interactions. Precedent demonstrates the value of fly studies for human cancer biology, due to conservation of molecular mechanisms. The long-term goal of our research is to use Drosophila to shed light on perhaps the most important question in cancer biology: Why Does Cancer Kill? In the previous funding cycle, we defined how Jun N-terminal Kinase (JNK) pathways activated by disruption of epithelial polarity induce a transcriptional program driving malignant proliferation. We also showed that malignant tumors induce cachexia-like wasting in adult hosts, and we identified the molecular mediator. In this cycle we will identify the molecular mechanism by which polarity loss activates oncogenic JNK signaling, and capitalize on our tumor-host paradigm to define how malignant tumors actually kill. Our first aim is to Determine the mechanism of TNFR activation during malignancy and wounding. Data indicate that JNK activation in tumors requires TNFR but not TNF, and we hypothesize that ligand-independent TNFR activation is controlled by its localization and/or post-translational modification. This may represent an underlying mechanism by which TNFRs can detect breaches in the epithelium in wounds as well as tumors. We will use genetics, structure-function analysis, and biochemistry to determine how disruption of epithelia triggers TNFR activation and initiates the wound-healing response which cannot be extinguished in tumors. Our second aim is to Define molecular and cellular mediators of tumor-host interactions. In addition to cachexia, our preliminary data extend tumor-host interactions to characterize the very early lethality induced by malignant transplants as well as edema possibly caused by renal failure. We hypothesize that tumor-produced factors independent of the wasting regulator drive `paraneoplastic' edema and host lethality, and will test the roles of candidate molecules and cell types in this signaling axis that are suggested by our preliminary data. Our third aim is to Identify determinants of host susceptibility to tumor lethality. We hypothesize that, as in infections, host factors play critical roles in influencing the course of malignant disease. We will identify tumor and host factors that either promote or restrict malignancy using an unbiased, forward genetic screen that exploits a new adult tumor model. Deciphering the mechanisms by which these factors ?particularly those that allow the host to tolerate a malignant tumor-- act in fly cancer could have a potentially transformative impact on understanding how human cancer patients respond, and what new interventions might be feasible.